SHORT THESIS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (PhD)

GENETIC ALTERATIONS IN CRANIOSYNOSTOSIS, GENOTYPE-

PHENOTYPE CORRELATIONS

by

Beáta Bessenyei

Supervisor: Prof. Dr. Éva Oláh

UNIVERSITY OF DEBRECEN DOCTORAL SHOOL OF CLINICAL MEDICINE

Debrecen, 2015

2

GENETIC ALTERATIONS IN CRANIOSYNOSTOSIS, GENOTYPE-PHENOTYPE CORRELATIONS

by Beáta Bessenyei, MSc

Supervisor: Prof. Dr. Éva Oláh, MD, PhD, DSc

Doctoral School of Clinical Medicine, University of Debrecen

Head of the Examination Committee: Prof. Dr. András Berta, MD, PhD, DSc

Members of the Examination Committee: Prof. Dr. István Raskó, PhD, DSc

Dr. Veronika Karcagi, PhD

The Examination takes place at the Department of Ophthalmology, Faculty of Medicine, University of Debrecen, 10 December, 2015., 12:00 p.m.

Head of the Defense Committee: Prof. Dr. András Berta, MD, PhD, DSc

Reviewers: Dr. Olga Török, MD, PhD

Dr. Emese Horváth, MD, PhD

Members of the Defense Committee: Prof. Dr. István Raskó, PhD, DSc

Dr. Veronika Karcagi, PhD

The PhD Defense takes place at the Lecture Hall of Bldg. A, Department of Internal Medicine, Faculty of Medicine, University of Debrecen, 10. december 2015., 13:30.

3

1. INTRODUCTION

Craniosynostoses, caused by the premature fusion of cranial sutures, are characterized by

abnormal skull shape. They occur in two main forms: in the isolated (non-syndromic) type,

craniosynostosis is not associated with other clinical signs, while in syndromic cases, facial

dysmorphism and limb anomalies can be observed as well.

Deformity of the skull is not only an aesthetic problem, but - depending on the type and

etiology of the disease -, it may lead to serious neurological, ophthalmological or respiratory

consequences requiring early surgical interventions. Therefore, early recognition and

treatment of craniosynostosis as well as determination of the subtypes are of great clinical

significance. Besides physical examination and applying imaging techniques (roentgen, CT),

identification of the genetic background is also important in the diagnostics of

craniosynostoses, in separation of isolated and syndromic forms and classification of

patients. Genetic alterations detected have a prognostic value and allow prenatal genetic

testing in further pregnancies; on the basis of the result, parents can decide upon the

outcome of the pregnancy.

Our knowledge of the clinical and genetic characteristics of patients with craniosynostosis in

Hungary is incomplete. Identification of the syndromic forms is a diagnostic challenge

because of their rarity and variable expressivity. A clinical center with experts experienced

in diagnostics and treatment of craniosynostoses, formed in 2005 in Hungary, can provide a

good opportunity for targeted and centralized medical care of these patients. From that time,

craniofacial reconstructions have been performed in the Department of Neurosurgery,

University of Debrecen, while the Department of Pediatrics provides the pre- and

postoperative treatment and genetic diagnostics of patients with the contribution of

neonatologists, pediatricians and clinical geneticists. This situation provides a great

4

opportunity to overview and summarize our experience in the clinical and genetic

characteristics of craniosynostosis and genotype-phenotype correlations.

2. LITERATURE REVIEW

Craniosynostoses, the premature fusion of cranial sutures, are characterized by the deformity

of the skull. It is one of the most frequent form of craniofacial disorders with an incidence of

1:2100-2500 births. Early fusion of the sutures usually begins prenatally, thus

craniosynostosis can be diagnosed mostly in newborns. In the isolated (non-syndromic)

forms (80-85% of patients), besides the cranial deformity and its potential consequences (e.

g., increased intracranial pressure, brain blood flow, respiratory distress, visual and hearing

impairment), no other specific clinical signs could be observed. In the contrary, in

syndromic forms the cranial abnormality often associates with further specific symptoms

including various facial dysmorhic signs and limb anomalies, however, there is an

overlapping between the phenotypes. More than 100 syndromes associated with

craniosynostosis are known nowadays.

2.1 Etiology of craniosynostosis

The isolated forms are mostly sporadic, their etiology is not entirely clarified yet. They are

considered to be of multifactorial origin: both the genetic background and environmental

factors may play an important role in their development. By epidemiological studies, the

main risk factors of isolated craniosynostoses are - among others- the male sex, high

(>4000g) or low (<1500g) birth weight, preterm (<37 week) birth, breech presentation,

plurality and advanced (>35 years) maternal age. Smoking, medication (e. g. antiepileptic

drugs) or alcohol use during pregnancy may also predispose to craniosynostosis. In contrast

5

with the isolated group, syndromic forms are mostly caused by genetic alterations, which

can be gene mutations or chromosomal aberrations.

2.1.1. Gene alterations

In about 25% of craniosynostoses, pathogenic mutations can be detected in genes playing

important role in the differentiation and proliferation of osteoblasts and ossification. The

most frequently involved genes in craniosynostosis are those encoding the fibroblast growth

factor receptors and the TWIST1 (twist family basic helix-loop-helix transcription factor 1)

transcription factor.

Fibroblast growth factor receptors (FGFR) and their genes

The four members of FGF receptor family (FGFR1-4) are transmembrane signal

transduction molecules with tyrosine kinase activity. They consist of an extracellular ligand

binding part with 3 immunoglobulin-like domains (IgI, IgII, IgIII), a transmembrane domain

and a splitted intracellular tyrosine kinase domain. The IgIII domain is located in the center

of the ligand binding part of the receptor, thus its amino acid sequence is crucial for the

specificity of ligand binding. The ligands of the receptors are the FGF molecules consisting

of 23 members, the ligand specificity of the receptors is different. They play an important

role in organogenesis, neurogenesis, wound heeling and endochondral or intramembranous

ossification.

In the development of craniosynostosis, FGFR1-3 genes are involved. Heterozygous,

activating mutations of FGFR2 can be identified in several syndromes such as Apert,

Crouzon or Pfeiffer syndromes. Most of these mutations are missense and located in the two

hotspot regions (exon IIIa and IIIc) of the gene. The p.Pro252Arg mutation in FGFR1 is

specific for Pfeiffer syndrome type 1, while in the FGFR3 gene, 2 mutations are known to

6

cause syndromic craniosynostosis (p.Pro250Arg, Muenke syndrome; p.Ala391Glu, Crouzon

syndrome with acanthosis nigricans). FGFR mutations cause gain of function, the receptors

become activated constantly. Enhanced signal transduction results in increased proliferation

and differentiation inducing intense ossification.

TWIST1 transcription factor and its gene

The TWIST1 molecule, a member of the basic helix-loop-helix transcription factor family,

has a crucial role in the development of the skeleton as a key molecule of mesenchymal cell

death. It has a positive or negative effect on the proliferation, differentiation and survival of

osteoblasts through the regulation of several signal transduction pathways. It regulates the

development of the cranial sutures indirectly through the BMP (bone morphogenetic

protein) and FGF pathways. Heterozygous, loss-of-function mutations in the coding exon of

TWIST1 are various: they can be missense, nonsense or frameshift mutations. In certain

cases, deletion of the whole gene or larger region can be found leading to microdeletion.

TWIST1 mutations cause craniosynostosis by haploinsufficiency.

The common craniosynostosis syndromes show autosomal dominant inheritance. In

syndromes with serious phenotype (e. g., Apert or Pfeiffer syndrome type 2), new mutations

are expected, while in milder syndromes (e.g., Muenke or Crouzon syndromes) familial

cases can also be observed.

2.1.2 Chromosomal abnormalities

Several chromosomal aberrations have been described in craniosynostosis, most of them are

sporadic. One part of the cytogenetic aberrations involves the 7p21.1 region containing the

TWIST1 gene, which translocation or deletion cause Saethre-Chotzen syndrome. Other

7

aberrations include terminal 9p, 11q23 or 22q11 deletions, 1p36 or 5q35 trisomies, etc. In

certain cases without chromosomal aberrations, various submicroscopic alterations may be

detected with multiplex ligation dependant probe amplification or array comparative

genome hybridization (arrayCGH). Chromosomal aberrations constitute about 15% of

genetically proved cases.

3. AIMS OF THE STUDY

1. The aim of our study was to investigate the clinical and genetic aspects of patients

with syndromic craniosynostosis referred to the Department of Pediatrics and

Department of Neurosurgery, and establish genotype-phenotype correlations.

2. We intended to use cytogenetic, molecular cytogenetic and molecular genetic

methods to detect genetic aberrations.

3. In the case of a new mutation, additional analyses were planned to use to confirm

its pathogenecity.

4. In the knowledge of the genetic aberration and the clinical signs, we studied

whether there is a correlation between the genotype and the phenotype, the

severity and the outcome of the surgical intervention.

5. In familial cases, examination of the proband was complemented with the

investigation of family members to enhance the better recognition of the rarer

syndromes.

6. On the basis of our results, we aimed to establish a genetic algorithm, which can

be effectively used in diagnostics.

8

7. Besides the investigation of the syndromic forms, we studied the frequency of

different types of isolated craniosynostosis, and the role of selected perinatal

factors in the development of the disorder.

4. PATIENTS AND METHODS

4.1.Patients

Two hundred patients with craniosynostosis were referred to the Department of

Neurosurgery and Department of Pediatrics, University of Debrecen between the years 2006

and 2012. Synostosis of the sutures was confirmed by either X-ray or computed

tomography, and the syndromic nature of the disease was established by a team of

pediatricians and clinical geneticists. Of the 200 enrolled patients, 198 were under 10 years

of age with a median age of 6 months (1 month-10 years), while only 2 patients were adults

(18 and 28 years old). Reconstructive operative procedures were performed in 195 patients.

In 24 patients, clinical signs other than the malformed skull suggested the syndromic forms

of the disease. Patients diagnosed with craniosynostosis syndromes during the seven years

of the study had the following types: Apert (n=5), Pfeiffer (n=5), Muenke (n=4), Crouzon

(n=2) and Saethre-Chotzen (n=1) syndromes. Multiple-suture craniosynostosis associated

with achondroplasia was found in one patient (n=1). Phenotypic features were not typical

for any particular syndrome in six patients.

4.2. Methods

To investigate the etiology of the disease genetic analyses were performed only in patients

with clinically identified or suspected syndromic craniosynostosis and their relatives

showing the clinical signs of a specific syndrome. Genetic testing included the molecular

9

analysis of the mutational hotspots of the FGFR1, FGFR2, FGFR3 and TWIST1 genes, G-

banded karyotyping and fluorescence in situ hybridization (FISH) analysis of TWIST1. If a

specific syndrome was identified (n=18), targeted analysis was performed, while in

suspected syndromic cases (n=6) all of the previously mentioned tests were applied. In one

patient arrayCGH analysis was also performed. After obtaining written informed consent,

blood samples were taken from 24 syndromic patients and 8 relatives. To assess the

pathogenecity of the novel mutation, DNA samples of 50 healthy, controll persons were

analyzed.

4.2.1. DNA extraction

To examine the mutational hotspots, genomic DNA was extracted from the patients’

peripheral blood using the QIAamp DNA Blood Mini kit (Qiagen, Germantown, MD)

according to the manufacturer’s instruction.

4.2.2. Mutational analysis of the FGFR1, 2, 3 and TWIST1 genes

Amplification of exon 7 of FGFR1, exons 8 and 10 of FGFR2, exons 7 and 10 of FGFR3,

and the coding region of exon 1 in TWIST1 was performed by polymerase chain reaction

(PCR) with previously published primer pairs. For DNA sequencing, the PCR products were

purified using the MinElute PCR Purification Kit (Qiagen, Germantown, MD, USA).

Purified PCR products were sequenced on the ABI 3100 sequencer (Applied Biosystems,

Foster City, CA, USA) with the Big Dye Terminator v3.1 cycle sequencing kit (Applied

Biosystems, Foster City, CA, USA). Electropherograms of the sequenced products were

compared to reference sequences.

10

4.2.3. Assessing the pathogenicity of the novel variant

To assess the pathogenicity of the novel variant, c.528C>G (p.Ser176Arg) in exon 1 of

TWIST1, SIFT (http://sift.jcvi.org) and Polyphen-2 (http://genetics.bwh.harvard.edu/pph2/)

predictions were performed. Alignment analysis of TWIST1 proteins was carried out using

Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/). The presence of the novel

variant was tested by restriction fragment length polymorphism (RFLP) analysis using

BspMI (Thermo Scientific, Waltham, MA, USA) restriction enzyme in 50 healthy

individuals.

4.2.4. Cytogenetic analysis

To detect the constitutional karyotype of the patients, conventional karyotyping was

performed. Blood samples (0.5 mL) of patients anticoagulated with Na-heparin were

cultured in medium (5 mL, Lymphochrome Medium, Lonza, Belgium) during 72 hours, at

37 degrees in a CO2 (5%) incubator. To interrupt cell division, colchicin (0.5 ug/ml, Sigma-

Aldrich, St. Louis, MO, USA) was added to the samples, and after incubation (1 hour),

hypotonization of the samples were performed with 0,075 M KCl solution. Fixation of the

samples was performed with methanol-acetic acid (3:1) solution. For chromosome banding,

treatment with trypsin (Sigma-Aldrich, St. Louis, MO, USA) was used followed by Giemsa

staining (Merck, Darmstadt, Germany). We evaluated 10 metaphases with Lucia Karyo

software (Lucia Cytogenetics, Czech Republik) in every patient. Determination of the

karyotype was based on the International System for Human Cytogenetic Nomenclature

(ISCN 2005 and 2009).

11

4.2.5. FISH analysis of TWIST1

FISH analysis was performed using cell suspension derived from chromosome preparation.

After dropping and aging (37 oC, one hour), slides were pretreated with the following

solutions: 2xSSC/0.5% NP40, 37 oC, 15 minutes; pepsin treatment, 37 oC, 5 minutes; 1x

PBS, room temperature, 5 minutes. After dehydration (70%-85%-100% ethanol, 2-2

minutes) and drying, 1.5 µl probe (Williams-Beuren/Saethre-Chotzen probe, Cytocell,

Cambridge, UK) was dropped onto the slide, and it was covered by a coverslip.

Codenaturation of the sample and the probe was performed at 76 oC for 3 minutes, and

hybridization was performed at 37 oC, overnight in a hybridization machine (Hybrite,

Abbott/Vysis, Des Plaines, IL, USA). We washed the unbounded probe in 50%

formamid/2X SSC solution at 42 °C, for 15 minutes, in 2X SSC solution at room

temperature for 10 minutes and in 2X SSC/ 0,1 % NP40 solution at room temperature for 5

minutes. After drying, cell nuclei were stained with 4’-6’ diamino-2-fenil-indol (DAPI,

Abbott/Vysis, Des Plaines, IL, USA). Evaluation of the samples was performed with Zeiss

Axioplan2 microscope (Carl Zeiss, Jena, Germany) and ISIS software (Metasystems,

Altlussheim, Germany). At least 15 metaphases were evaluated in every patient.

4.2.6. ArrayCGH analysis

ArrayCGH analysis was performed in a patient with multiple-suture craniosynostosis and

achondroplasia at the Department of Clinical Genetics, Academic Medical Centrum,

Amsterdam, Netherlands with Agilent 180K oligo-array, Amadid 023363 (Agilent

Technologies, Inc., Santa Clara, CA, USA). Standard methods were used for labelling and

hybridization, array profiles were evaluated by Agilent software.

12

4.2.7. Statistical analysis

To assess the potential risk factors for craniosynostosis, selected perinatal data (sex,

plurality, gestational age at birth and birth weight) of 142 non-syndromic patients were

compared with the general Hungarian live births data (average data of 2005-2010 years)

provided by the Hungarian Central Statistical Office. Statistical analyses included Chi-

square and Fisher’s exact tests performed by IBM SPSS 20 (IBM Corporation, Armonk,

New York). The results were considered to be significant at a < 0.05 significance level.

5. RESULTS

Two hundred patients with craniosynostosis were enrolled in the study. On the basis of

detailed clinical assessment, the condition proved to be isolated (nonsyndromic) in 176 of

200 (88%) patients while 24 cases (12%) were syndromic.

5.1. Clinical symptoms of syndromic patients

Clinical signs in patients with Apert syndrome included brachycephaly due to bicoronal

synostosis, high and broad forehead, midface hypoplasia, depressed nasal bridge, convex

nasal ridge, hypertelorism and syndactyly of fingers and toes. In the majority of patients,

syndactyly of all toes on both feet and syndactyly of the 2/3/4/5 fingers on both hands were

observed (Patients 1-4). In Patient 5, complete syndactyly of all fingers and all toes could be

seen. Cleft palate was observed in one patient.

Both patients with Crouzon syndrome showed brachycephaly, proptosis, flat nasal bridges,

mandibular prognathism and low-set ears. In one of them acanthosis nigricans was also

present.

13

Three of the patients with Pfeiffer syndrome had severe cranial malformation (cloverleaf or

“Kleeblattschädel” skull), hydrocephalus, extreme proptosis, low-set ears and short, small

noses. Broad great toes were observed in all of these patients, while the thumbs were broad

only in two of them. Limited extension of the elbows was also noticed in two patients. In a

patient, medially deviated great toes, the characteristic feature of Pfeiffer syndrome, were

observed and were associated with partial 2/3 syndactyly on the feet. Based on the

phenotypic signs, the latter three patients were considered to have Pfeiffer syndrome type 2.

In two further patients, Pfeiffer syndrome type 1 was diagnosed based on craniofacial

anomalies milder than those in patients with type 2. One of them had brachycephaly, a high

and broad forehead, exophthalmos, broad thumbs, 3/4 syndactyly of the fingers, broad great

toes and 2/3/4 syndactyly on the toes. The other patient had brachy- and acrocephaly, high

forehead, wide and depressed nasal bridge, hypertelorism, long philtrum, thin lips, low set

ears with overfolded helices. Limb anomalies included broad thumbs, broad great toes with

varus deformity, complete 3/4 cutaneous syndactyly of the right foot and 2/3/4 partial

cutaneous syndactyly of the left. Detailed clinical and genetic investigations of the other

patient and her 4 family members were performed.

All of the Muenke syndrome patients had brachycephaly resulting from the synostosis of the

coronal sutures, high and flat foreheads, hypertelorism and almond-shaped eyes. High-

arched palate was seen in two of them. Limb abnormalities were variable, including broad

thumbs and great toes, clinodactyly of the fifth fingers and capitate-hamate fusion. The

mothers of three patients showed similar phenotypes to those of their children, but their

clinical signs seemed to be milder.

Typical clinical signs of Saethre-Chotzen syndrome were observed in a patient, who had

brachycephaly, an asymmetric flat face, long and deviated nose, thin lips, ptosis of the right

14

eyelid, shallow orbits, and broad and bifid great toes. The mother of the patient also had

brachycephaly, a long nose, thin lips and bifid great toes; additionally, her thumbs were

broad.

Multiple-suture craniosynostosis associated with rhizomelic shortening of the limbs and

trident hands, the typical clinical signs of achondroplasia were found in a patient.

In six patients, additional clinical features associated with craniosynostosis included various

facial dysmorphic signs and limb alterations suggesting the syndromic form of

craniosynostosis; however, the phenotypes of these patients were not typical for any

particular syndrome.

5.2. Genetic alterations in syndromic patients

The Apert and Pfeiffer syndromes could most commonly be observed in the syndromic

group. Genetic abnormalities were detected in 75% (18/24) of the syndromic patients and in

8 relatives showing the clinical signs of a specific syndrome. All mutations were

heterozygouos.

In 4 patients with Apert syndrome, we identified the c.758C>G (p.Pro253Arg) mutation in

FGFR2, while in one patient the c.755C>G (p.Ser252Trp) mutation of the same gene could

be detected. Both mutations are specific for this syndrome.

In 4 patients with Pfeiffer syndrome, the FGFR2 gene was involved: in three cases various

changes of the 342. amino acid were detected (c.1024T>C, p.Cys342Arg és c.1025G>C,

p.Cys342Ser), while in one patient a splicing mutation (c.940-1G>A) could be identified.

The rare c.755C>G p.Pro252Arg mutation of FGFR1 was detected in the patient with

Pfeiffer syndrome type 1 characterized by mild symptoms. Detailed clinical genetic

15

examination of the family was performed, and the mutation was proved in further 4 family

members.

In one of the patients with Crouzon syndrome the c.833G>T (p.Cys278Phe) mutation of

FGFR2 was detected, in the other patient with acanthosis nigricans, the c.1172C>A

(p.Ala391Glu) mutation in the FGFR3 gene was found.

The mutation specific for Muenke syndrome, the c.749C>G (p.Pro250Arg) in FGFR3, could

be detected in 4 patients, three of them inherited it from their mothers.

In the patient with achondroplasia and multiple-suture craniosynostosis, the achondroplasia-

specific c.1138G>A (p.Gly380Arg) mutation was found. Any other genetic alteration which

may have been related to the associating craniosynostosis, could not be identified by

sequencing of the hotspot regions of FGFR1-3, G-bandig karyotyping, FISH and arrayCGH.

In addition to 10 different, known mutations detected in FGFR1-3, one previously

undescribed missense mutation, the c.528C>G (p.Ser176Arg), was found in the TWIST1

gene of a patient with Saethre-Chotzen syndrome.

5.3. Assessing the pathogenecity of the new mutation p. Ser176Arg

The pathogenicity of the new mutation, c.528C>G (p.Ser176Arg), is supported by the

following arguments:

1. the affected residue is phylogenetically highly conserved in human, mouse,

zebrafish and xenopus;

2. according to SIFT and PolyPhen-2 predictions, this missense alteration was

predicted to be damaging and probably damaging, respectively;

3. RFLP analysis of exon 1 of TWIST1 showed that this mutation was not present in

50 healthy individuals representing 100 alleles.

16

4. The mother, who had very similar phenotype also had the novel c.528C>G

(p.Ser176Arg) mutation.

5.4. The role of selected perinatal risk factors in non-syndromic craniosynostoses

In the isolated group, the most frequently fused suture was the sagittal one (120/176; 68%)

followed by the coronal (26/176; 15%), metopic (18/176; 10%) and lambdoid sutures

(7/176; 4%) in this order. In 5 patients (3%), more than one suture was involved. The male-

to-female ratio was 2.2:1, showing a male predominance.

Perinatal data of 142 nonsyndromic patients were compared to the general Hungarian live

birth data to assess their possible associations with craniosynostosis. Male sex (P< 0.001),

twin gestation (P< 0.001) and very low (< 1500 g) birth weight (P< 0.001) proved to be risk

factors for nonsyndromic craniosynostosis. Being male (P< 0.001) and twin gestation

(P=0.001) seemed to be associated with sagittal synostosis, while very low birth weight was

a predictor for coronal (P<0.001) synostosis. No association between metopic synostosis

and any risk factor analyzed in the study could be confirmed. Because of the small number

of patients, no statistical analysis could be performed on lambdoid synostosis.

6. DISCUSSION

Our study is the first nation-wide investigation providing clinical and genetic information on

craniosynostoses. Similarly to previously published studies, nonsyndromic patients

represented the majority of all cases, while syndromic forms constituted only the 12% of

craniosynostoses.

17

6.1. Genetic alterations in syndromic craniosynostoses

Genetic abnormalities were detected in syndromic patients with clinical signs characteristic

for a specific syndrome.

6.1.1. Genotype-phenotype correlations

Apert syndrome is one of the most serious craniosynostosis syndromes with specific clinical

signs. According to the literature, the specific mutations p.Ser252Trp and p.Pro253Arg of

FGFR2 can be detected in about 99% of patients with Apert syndrome, while Alu-element

insertions in FGFR2 are a rare cause of this condition. In our study, all patients with Apert

syndrome carried one of the two specific mutations. While the p.Ser252Trp mutation could

be detected in only one patient in our study, other authors found this mutation with a higher

frequency. Regarding the clinical features, cleft palate has been reported to be more

common with the p.Ser252Trp mutation, while syndactyly was found to be more severe in

patients with the p.Pro253Arg mutation. Our result supports this observation: the only

patient, who had cleft palate, carried the p.Ser252Trp mutation.

Pfeiffer syndrome is known to be clinically and genetically heterogenous disorder. The two

ends of the phenotypical spectrum are represented by the very mild, almost unrecognizable

cranial deformity on one hand and the severe cloverleaf skull on the other. The mutations

eliminating the cysteine 342 residue in FGFR2 have a different phenotypic impact

depending on the type of amino acid exchange. The conversion of cysteine to phenylalanine

or tyrosine mainly results in Crouzon syndrome, whereas arginine substitution preferentially

causes Pfeiffer syndrome with severe cranial manifestation and poor prognosis. In our

cohort, two patients with Pfeiffer syndrome harbored the p.Cys342Arg mutation; both of

them had cloverleaf skulls requiring two and three reconstructive surgical interventions,

18

hydrocephalus, severe exophthalmos, and respiratory and auditory problems. In addition,

exchange of the same amino acid residue for serine caused very similar, severe phenotype in

the third patient with Pfeiffer syndrome type 2. The mild form of Pfeiffer syndrome, the

type 1 was observed in two patients. In one of them, splicing mutation of FGFR2 (c.940-

1G>A) was identified, while the other patient had the p.Pro252Arg mutation in FGFR1.

In 1994 Muenke et al. identified the FGFR1 gene on chromosomal region 8p11 to be

associated with Pfeiffer syndrome. A specific mutation, p.Pro252Arg, located in the IgII-III

linker region of FGFR1, has been reported to cause mild symptoms making the diagnosis

rather difficult in many cases. FGFR1 p.Pro252Arg mutation without craniosynostosis has

been reported on two occasions in the literature providing a further evidence of variable

expressivity of the syndome. Phenotypic features in the family with the mutation above

showed high variability ranging from apparently normal skull and limbs to characteristic

brachycephaly and digital anomalies. Typical features of the syndrome appeared only in the

third generation, suggesting that this condition is underdiagnosed in many cases. In the case

of FGFR1 p.Pro252Arg mutation, a variable expressivity can b e expected which in mild

cases makes the diagnosis difficult. The case of this family emphasizes the significance of

detailed physical examination of not only the proband but the family members as well,

because the specific symptoms of the syndrome may occasionally appear only after several

generations.

Crouzon syndrome is usually caused by FGFR2 mutations. Association of the syndrome

with acanthosis nigricans is very rare and considered to be a separate entity called

Crouzonodermoskeletal syndrome because of its specific phenotype and genotype. The

specific alteration of the syndrome is the c.1172C>A (p.Ala391Glu) mutation of FGFR3. In

contrast to the common FGFR1 and 2 mutations located in the extracellular part of the

19

receptor, this mutation is found in the transmembrane region of FGFR3 enhancing

signalization by the stabilization of receptor dimers. In one patient with Crouzon syndrome,

a common mutation of the syndrome, c.833G>T, p.Cys278Phe in FGFR2, was found, while

in the other patient acanthosis nigricans was also present and the former specific mutation

was detected. Because of the former observations, besides cranial deformity and facial

dysmorphism we must pay special attention to the skin, but limb anomalies are not expected.

Muenke syndrome is characterized by the synostosis of the coronal suture, very mild facial

dysmorphism, limb anomalies which can be seen sometimes only by X-ray imaging and the

presence of the specific FGFR3 p.Pro250Arg mutation. The syndrome was identified in 4

patients, three probands inherited the mutation from their mothers. Recent international

studies recommend the mutational analysis in every patient with isolated craniosynostosis

involving the coronal suture in order to detect the very mild, almost unrecognizable

assocating synostosis and facial dysmorphy. In a patient with Saethre-Chotzen syndrome we

succeeded to identify a novel missense mutation, p.Ser176Arg, in TWIST1. TWIST1 is a

basic helix-loop-helix transcription factor involved in a variety of signal transduction

pathways in tissues of mesodermal origin. It has a crucial role in the migration and

differentiation of cranial neural crest cells during cranial development. The novel alteration

is located between the highly conserved bHLH and Twist box domains. Although this linker

region is evolutionarily conserved across a wide range of vertebrate species, its functional

importance is still unclear. Pathogenecity of the variant is supported by the conservity of the

176. amino acid of the human TWIST1, the probable damaging effect of serine-arginine

exchange, the lack of the mutation in the healthy control population examined, and its

presence in the mother showing the clinical signs of the syndrome.

20

6.1.2. Association of craniosynostosis with achondroplasia

Association of achondroplasia, as the most common form of chondrodysplasias, with

craniosynostosis is extremely rare, only 3 cases have been reported before our publication.

In our patient, acrocephaly and left posterior plagiocephaly was observed due to multiple-

suture craniosynostosis. The achondroplasia-specific mutation could be detected, but a

second pathogenic mutation which may have lead to the simultaneously existing

craniosynostosis could not be identified in the hot-spot regions of FGFR1, 2, 3 and TWIST1.

Cytogenetic and array CGH analysis also gave normal result. Our case and the reported ones

suggest that this combined phenotype may be related to a variable expressivity of the

common mutation of achondroplasia. It can be suspected, that modifier gene(s) may exist in

the genome altering the phenotypic outcome of the achondroplasia-specific mutation. In

addition, epigenetic or environmental influences may also modify the manifestation of the

disease.

6.2 Characteristics of isolated craniosynostoses – predisposing factors

Similarly to previously published studies, nonsyndromic patients represented the majority of

all cases with craniosynostosis, and the most frequently involved suture was found to be the

sagittal one, followed by the coronal, metopic and lambdoid sutures. The frequency of

sagittal synostosis was 68%, higher than in other populations. Assessing the relationship

between potential risk factors and craniosynostosis, it was found - in accordance with

previous studies - that sagittal synostosis was frequently associated with male sex and twin

gestation. Several studies confirmed the role of maternal smoking as a risk factor for

craniosynostosis, especially the sagittal type. According to data from the World Health

Organization (2011), the estimated prevalence of daily smoking in Hungarian females is

21

26%, which is much higher than in many other countries (United Kingdom, 14%; United

States, 13%; Australia, 15%). Although smoking behavior of mothers was not studied in our

patient cohort, we can speculate that smoking can be one of the potential causes of the

higher involvement of the sagittal suture in Hungarian patients with craniosynostosis.

6.3. Diagnostic aspects and genetic algorithm

Detailed clinical genetic examination has an important role in not only the recognition the

different forms of the disease, but also has a prognostic significance and allows early

surgical intervention.

On the basis of our experience and data in the international studies, we established a

molecular diagnostic algorithm for the genetic examination of craniosynostosis in the

country. If the clinical signs are specific for a certain syndrome, targeted analyses should be

performed in a given order to maximize the diagnostic effectivity. If the symptoms propose

the syndromic form of craniosynostosis, but they are not specific for a certain syndrome,

conventional cytogenetic analysis is recommended at first, followed by the mutational

analysis of hotspot regions of the responsible genes.

As the etiology of the isolated group is very heterogeneous and probably is of multifactorial

origin, genetic tests are proposed only in certain cases (e.g., involvement of the coronal

suture, multiple-suture synostosis, familial cases). From a clinical point of view, it is

important, that the knowledge of the genetic background may call our attention to the

expected associating clinical signs and anticipate the probable required number of surgical

interventions. In genetically determined forms, the probability of refusion, pansynostosis

and progression is greater. Reconstruction of the cloverleaf skull in our patients with Apert

syndrome recquired 2 or 3 surgical intervetions, while the correction of syndactyly also

22

needed several operations. Besides the treatment of craniofacial alterations, increased

intracranial pressure (hydrocephalus), extreme proptosis and respiratory problems also

required surgical interventions in the Pfeiffer syndrome of severe type 2 and 3.

Crouzon, Muenke and Pfeiffer syndrome type 1 are usually characterized by mild cranial

deformity. As a result of this, three patients with Muenke syndrome in our cohort did not

need any reconstruction.

Additional significance of the genetic diagnosis is the possibility of targeted prenatal testing

in familial cases. On the basis of the result, couples can decide the outcome of the

pregnancy. As a result of our contribution, three healthy babies have been born so far.

New results of the dissertation

1. We performed comprehensive clinical and genetic investigation of patients with

craniosynostosis at first in the country. Taking the advantage of the leader role of

the Department of Neurosurgery in craniofacial surgery, we performed a national

center for the clinical and genetic examination of patients with craniosynostosis.

2. We achieved expansive genetic examination of syndromic craniosynostoses and

established a diagnostic algorithm for effective and economical testing.

3. In 75% of syndromic cases, genetic alterations could be detected.

4. We identified a novel mutation (p.Ser176Arg) in the TWIST1 gene of a patient

with Saethre-Chotzen syndrome, and supported its pathogenecity with additional

tests.

5. By the investigation of genotype-phenotype correlations, several observation were

made:

23

- We confirmed the correlation between the type of the mutation and the

phenotype in Apert, Crouzon and Pfeiffer syndromes.

- We performed genotype–phenotype analyses in the family of a patient with

Pfeiffer syndrome type 1 and p.Pro252Arg mutation and confirmed, that this

mutation is associated with variable expressivity.

- We noticed, that specific signs of the syndrome appear only after several

generations, which call attention to detailed family studies.

6. We confirmed the previous observation, that the rare association of

achondroplasia and multiple-suture craniosynostosis is the consequence of the

p.Gly380Arg mutation of FGFR3, which supports the role of this mutation both in

endochondral and intramembranous ossification.

7. On the basis of statistical analyses, we stated that male sex, twin gestation and

very low birth weight are risk factors for non-syndromic craniosynostosis.

8. We observed, that the involvement of sagittal synostosis in isolated patients is

significantly higher in the Hungarian population than in other countries, which

underlines the significance of further investigation of risk factors.

24

7. SUMMARY

Craniosynostosis, the premature fusion of cranial sutures, is a clinically and etiologically

heterogeneous group of disorders. Both environmental and genetic factors play an important

role in the development of isolated (non-syndromic) forms found in the majority of patients,

while syndromic cases are frequently caused by gene mutations or more rarely chromosomal

aberrations. The aim of our work was to study the clinical and genetic characteristics of

patients with craniosynostosis syndromes and to assess the genotype-phenotype correlations.

In isolated craniosynostoses, the pathogenetic role of infant sex, birth weight, gestational

age and plurality was retrospectively studied.

Using various methods, pathogenic mutations were identified in 75% of syndromic patients.

In five cases, the mutations proved to be familial. In the FGFR1, 2, 3 and TWIST1 genes, 11

different mutations were detected. No chromosomal aberrations could be identified. In a

patient with Saethre-Chotzen syndrome, a novel pathogenic mutation, p.Ser176Arg, in the

TWIST1 gene was found. In a family with Pfeiffer syndrome due to FGFR1 p.Pro252Arg

mutation, we observed a variable expressivity with specific signs of the syndrome appearing

only in successive generations. In a patient with simultaneous achondroplasia and multiple-

suture craniosynostosis, the achondroplasia-specific mutation, p.Gly380Arg, was confirmed

without any further genetic abnormality suggesting that this mutation has an influence on

both endochondral and intramembranous bone formation. Based on our experiments, we

developed an algorithm for genetic diagnosis of syndromic craniosynostosis. A relationship

between selected perinatal risk factors (male sex, twin gestation and low birth weight) and

isolated craniosynostosis was found. Clarification of the genetic background of

craniosynostosis and establishment of genotype-phenotype correlations are of important

diagnostic and prognostic significance offering the possibility for targeted prenatal testing.

25

26

27

28

29